Exclusive Peripheral Overexpression of 5-HT5A Receptors is Involved in 5‑HT‑Induced Cardiac Sympatho-Inhibition in Type 1 Diabetic Rats


 Background: In normoglycaemic pithed rats, cardiac sympathetic control is modulated by serotonin (5-hydroxytryptamine; 5‑HT), which inhibits the cardioaccelerator sympathetic outflow via the activation of 5-HT 1B , 5-HT 1D and 5-HT 5A receptors. Notwithstanding, type 1 diabetes impairs the functionality of the cardiac sympathetic innervation and leads to cardiovascular complications including cardiac autonomic neuropathy. On this basis, the present study investigated whether the influence of 5-HT on cardiac noradrenergic neurotransmission is altered in type 1 diabetic rats, by analysing the profile of the 5-HT receptors involved and their peripheral expression.

Methods: Type 1 diabetes was induced in male Wistar rats with a single injection of streptozotocin (50 mg/kg, i.p.). Four weeks later, the rats were anaesthetized, pithed and prepared for producing tachycardic responses by either electrical preganglionic stimulation (C 7 ‑T 1 ) of the cardioaccelerator sympathetic outflow or i.v. bolus injections of exogenous noradrenaline. Immunohistochemistry analyses were performed to study the expression of 5‑HT 1B , 5-HT 1D and 5-HT 5A receptors in the stellate (sympathetic) ganglion obtained from normoglycaemic and diabetic rats.

Results: The increases in heart rate evoked by both cardiac sympathetic stimulation and exogenous noradrenaline were not modified after saline in diabetic rats. Moreover, i.v. continuous infusions of 5‑HT induced a cardiac sympatho-inhibition that was mimicked by the 5‑HT 1/5A receptor agonist 5‑carboxamidotryptamine, but not by the agonists indorenate (5-HT 1A ), CP 93,129 (5‑HT 1B ), PNU 142633 (5-HT 1D ), or LY344864 (5‑HT 1F ) in the diabetic group. In contrast, the above agonists at 5-HT 1B , 5-HT 1D and 5-HT 1/5A receptors mimicked 5-HT-induced sympatho-inhibition in normoglycaemic rats. In diabetic animals, i.v. administration of SB 699551 (1 mg/kg; 5‑HT 5A receptor antagonist) abolished 5‑CT-induced cardiac sympatho-inhibition. Finally, the immunohistochemistry analysis in the stellate ganglion showed that, as compared to normoglycaemic rats, in diabetic rats (P<0.05): (i) the expression of 5-HT 1B receptors was slightly higher, whereas that of 5-HT 1D receptors was slightly lower; and (ii) there was a clear overexpression of 5-HT 5A receptors.

Conclusions: Taken together, these results show the prominent role of the peripheral overexpression of prejunctional 5-HT 5A receptors in the inhibition of the cardiac sympathetic drive in type 1 diabetic rats. These findings may represent a new pharmacological strategy for the treatment of diabetes-related cardiac abnormalities.


Background
The main causes of high morbidity and mortality in patients suffering from type 1 diabetes (T1D) 2 continue to be cardiovascular complications, which are ten-fold increased in patients with T1D 3 than in the general population [1,2]. Despite the notable improvement in life expectancy, patients 4 with type 1 diabetes confront this increased cardiovascular risk, as it is silent in the at early stages 5 of the disease (underestimated event), and due to the longer duration of the diabetic state 6 compared to type 2 diabetes [3][4][5]. One of the triggers for cardiovascular disorders associated 7 with diabetes is the sympathetic nervous system disorder at the cardiac level [6,7]. Indeed, an 8 increase in sympathetic (i.e. noradrenergic) heart tone has been established to contribute to 9 myocardial damage and therefore to cardiac disorders in patients with T1D [4,[8][9][10], as well as in 10 animal models of T1D [11,12]. The impairment of cardiac sympathetic neurotransmission, 11 among others, is strongly involved in the development of cardiac autonomic neuropathy (CAN), 12 a microvascular complication leading to dysfunctional heart rhythm and vascular dynamics [4,9,13 10]. In view of all this, the down-regulation of cardiac noradrenergic neurotransmission may be a 14 potential therapeutic approach to manage these cardiovascular disorders in T1D. 15 Within this context, 5-HT stands out as a neurohormone capable of regulating the 16 sympathetic nervous system; certainly, the serotonergic system has been shown to modulate 17 peripheral sympathetic neurotransmission at both the vascular and cardiac levels [13][14][15][16][17][18][19]. With 18 respect to perivascular sympathetic innervation in pithed rats, our research team has already 19 demonstrated that 5-HT induces sympatho-inhibition in the vasopressor discharge, mainly 20 through activation of prejunctional 5-HT1A and 5-HT1D receptors [13,14]. However, the 21 induction of T1D in rats altered the pharmacological profile of the vascular sympatho-inhibitory 22 5-HT receptors (which depended on the duration of diabetes). Specifically, our results revealed 23 the exclusive role of prejunctional 5-HT1A receptors in 28-day (short-term) diabetic rats [20], and 24 J.A. García-Pedraza et al., 2020 5 the additional role (besides 5-HT1A receptors) of prejunctional 5-HT2A receptors in 56-day (long-1 term) diabetic rats [21]. On the other hand, regarding cardiac sympathetic control, 5-HT 2 decreases the sympathetic chronotropic outflow in normoglycaemic-pithed rats through 3 prejunctional 5-HT1B, 5-HT1D and 5-HT5A receptors [15][16][17]22]. Despite the key participation of  Based on the above findings, the objective of the present study was to examine the role of the 9 5-HT system in the cardiac sympathetic modulation during T1D. In particular, the impact of

16
Animals and general methods 17 A total of 120 male Wistar rats (250-300 g) was used. The animals were maintained at a 12/12-h 18 light/dark cycle (light beginning at 7 a.m.) and housed in a special room at constant temperature 19 (22 ± 2°C) and humidity (50%), with food and water ad libitum. 20 The rats (n=120) were initially divided into 2 main sets, namely, set 1 (diabetic rats, 21 n=75) and set 2 (normoglycaemic rats, n=45). In set 1, diabetes was induced by an i.p. injection 22 of streptozotocin (STZ, 50 mg/kg) which was dissolved in citrate buffer (pH=4.5); whereas set 2 23 received an i.p. injection of vehicle (citrate buffer, 1 ml/kg) as previously reported [28,29]. Both 24 J.A. García-Pedraza et al., 2020 6 sets were kept for 4 weeks. Blood glucose levels and body weight were determined before and 1 weekly after administration of STZ or vehicle. It is noteworthy that in set 1 all rats were 2 considered for experimentation as they displayed elevated blood glucose levels (i.e. >300 mg/dl) 3 at all time-points. In all animals, blood glucose levels were measured using Accu-Chek strips and 4 a glucometer (Accu-Chek Performa, Roche Diagnostics, Mannheim, Germany); and the body  After 4 weeks of treatment with STZ or vehicle, all animals were initially anaesthetized 8 with sodium pentobarbital (60 mg/kg, i.p.). The adequacy of anaesthesia was judged by the 9 absence of ocular reflexes, a negative tail flick test and corporal relaxation, amongst others. At 10 this point, 5 normoglycaemic and 5 diabetic rats were used for analysing the expression of 5-HT 11 receptors by immunohistochemistry (see section Immunohistochemistry analysis of 5-HT 12 receptors). The rest of the rats (n=110), consisting of set 1 (n=70) and set 2 (n=40), were used for 13 pharmacological tests in pithed animals (see Fig. 1).
14 After cannulation of the trachea, the rats were pithed by inserting a stainless-steel rod 15 through the ocular orbit and foramen magnum into the vertebral foramen. Immediately 16 afterwards, the animals were artificially ventilated with room air using an Ugo Basile pump (56 17 strokes/min and a stroke volume of 20 ml/kg; Ugo Basile Srl, Comerio, VA, Italy) [22,29]. After neurotransmission (n=100, i.e. 60 rats from set 1 and 40 rats from set 2); or (ii) i.v. bolus 5 injections of exogenous noradrenaline (n=10 rats from set 1).

6
The tachycardic stimulus-response (S-R) curves and the dose-response (D-R) curves 7 elicited by, respectively, preganglionic sympathetic stimulation and exogenous noradrenaline 8 were completed in about 30 minutes, with no changes in the baseline values of resting heart rate 9 or blood pressure. The sympathetic tachycardic stimulation (0.03-3 Hz) and the i.v. dosing with 10 exogenous noradrenaline (0.03-3 μg/kg) were given using a sequential schedule in 0.5 log unit 11 increments, at intervals of 5 minutes. The body temperature of each pithed rat was maintained at 12 37°C by a lamp and monitored with a rectal thermometer.

13
Experimental procedures 14 After a stable haemodynamic condition for at least 30 min, baseline values of heart rate and 15 diastolic blood pressure (a more accurate indicator of peripheral vascular resistance) were 16 determined. Then, the two sets of rats underwent the following experimental protocols (Fig. 1). 17 Protocol I. Electrical stimulation of the cardiac sympathetic innervation 18 In this protocol (n=100, i.e. 60 from set 1 and 40 from set 2; see Fig. 1), the pithing rod was 19 replaced by an electrode enamelled except for 1 cm length 7 cm from the tip, so that the 20 uncovered segment was situated at C7-T1 of the spinal cord. This procedure allowed selective   When heart rate had returned to baseline levels, the next frequency was applied; this procedure 9 was systematically performed until the S-R curve was completed. Subsequently, this set of 10 animals was divided into two clusters (normoglycaemic and diabetic rats; see Fig. 1). 11 The first cluster (normoglycaemic; n=40) was used here for confirming previous results  The second cluster (diabetic; n=60) was divided into 3 groups (n=40, 10 and 10). The first 22 group (n=40), subdivided into 8 subgroups (n=5 each), received i.v. continuous infusions of 23 saline or the 5-HT receptor agonists exactly as described above for the first cluster (see Fig. 1).

24
Ten min after starting each infusion, a S-R curve was elicited again as described above. The 1 second group (n=10), subdivided into 2 subgroups (n=5 each), received i.v. injections of saline 2 (1 ml/kg), or SB 699551 (5-HT5A receptor antagonist, 1 mg/kg). Ten min later, a S-R curve was 3 elicited again to analyse the effects of these compounds per se on the sympathetic tachycardic 4 responses. And the third group (n=10), subdivided into 2 subgroups (n=5 each), received i.v. 5 bolus injections of saline (1 ml/kg), or SB 699551 (1 mg/kg). Ten min later, these subgroups 6 received an i.v. continuous infusion of 5-CT (0.1 μg/kg.min). After 10 min, a S-R curve was 7 elicited again as described above. The last set of diabetic rats (n=10) was prepared as described above, but the pithing rod was left 10 and the administration of both gallamine and desipramine was omitted. Then, the rats received  Subsequently, this set, divided into two groups (n=5 each), received i.v. continuous infusions 14 of saline (0.02 ml/min), or 5-CT (0.1 μg/kg.min). Ten min later, a D-R curve was elicited again 15 during the infusion of the above compounds. 16 Other procedures applying to protocols I and II 17 The doses of all compounds used in the pharmacological tests of the present study have 18 previously been reported [16,17,22]. The 5-HT receptor agonists and saline were continuously 19 infused at a rate 0.02 ml/min by a WPI model sp100i pump (World Precision Instruments Inc., 20 Sarasota, FL, U.S.A.). The intervals between the different stimulation frequencies or 21 noradrenaline doses depended on the duration of the tachycardic responses (5-10 min), as we 22 waited until heart rate had returned to baseline values. 1 For this purpose, as previously pointed out, 5 normoglycaemic and 5 diabetic rats were used; 2 these animals were sacrificed with an overdose of sodium pentobarbital (180 mg/kg; i.p.). Then, 3 the upper mediastinal block was extracted by careful dissection, together with the content of the 4 clavicular space, which was fixed overnight by immersion in 4% paraformaldehyde in phosphate in TBS) over night at 4° C in a humidity chamber. After washing, the slides were incubated for 15 2 hours with goat anti-rabbit IgG Alexa Fluor ® 488 (Abcam ® , diluted at 1:600 in TBS). 16 Afterwards, the sections were washed again with TBS followed by immersion in Mayer's 17 haematoxylin to visualize the nuclei and identify anatomical structures. Slides were mounted 18 with fluoromount aqueous mounting medium (Sigma ® ). The fluorescence visualization in slides 19 was obtained by a confocal microscope TCSSP2 (Leica ® ). Z-stack of 5 x 1 μm sections were 20 captured from the stellate ganglion. From 40x high-resolution micrographs, fluorescence 21 intensity was determined as arbitrary units of grey level with ImageJ free viewer software (NIH). 23 All data in the text, tables and figures are presented as the mean ± SEM. Initially, the difference 1 between the changes in blood glucose and gain of body weight in normoglycaemic and diabetic 2 animals was evaluated using Mann-Whitney U test. The peak changes in heart rate by cardiac 3 sympathetic stimulation or exogenous noradrenaline in the saline-and 5-HT receptor agonists- 4 infused rats were determined. The difference in the values of heart rate and diastolic blood 5 pressure within one (sub)group of rats before and during the infusions of saline or the 5-HT 6 receptor agonists, as well as the immunohistochemistry data, were evaluated with paired 7 Student's t-test. Moreover, the difference between the changes in heart rate within one subgroup 8 of animals was evaluated with the Student-Newman-Keuls test, once a two-way repeated 9 measures ANOVA (randomized block design) had revealed that the samples represented different 10 populations. Additionally, the difference between measured baseline haemodynamic variables in 11 diabetic and normoglycaemic pithed rats was evaluated using a one-way ANOVA, followed by

9
Systemic haemodynamic parameters 10 As compared to the normoglycaemic (vehicle-injected) rats, STZ administration elicited a 11 marked increase in blood glucose concentration (mg/dl) and a decrease in body weight (g).
12 Table 1 shows the mean values of these parameters before (initial) and 4 weeks after the 13 administration of STZ or its vehicle in rats. 14 Certainly, at the very beginning of the experiments, we used two different body weights 15 for the administration of vehicle or STZ because the tachycardic responses induced by electrical 16 stimulation are optimally produced in animals weighing around 290 g [28,29]. Thereupon, as the 17 i.p. administration of STZ resulted in a marked loss of body weight, we decided to use animals of 18 approximately 400 g so that, after 4 weeks, this body weight decreased to approximately 290 g.

19
In contrast, for the i.p. administration of vehicle we used animals of approximately 200 g in order 20 that, after 4 weeks, this body weight increased to around 290 g (approximating that in diabetic 21 rats; see Table 1). 22 Furthermore, the baseline values of heart rate and diastolic blood pressure in the 23 normoglycaemic rats (n=40) were 249±7 beats/min and 50±2 mmHg, respectively; whereas in the 24 J.A. García-Pedraza et al., 2020 13 diabetic rats (n=70) these values were 219±7 beats/min and 48±1 mmHg, respectively. When 1 comparing these values between the normoglycaemic and diabetic rats, diastolic blood pressure 2 values did not significantly differ (P>0.05), but the heart rate values were significantly lower in 3 diabetic rats (P<0.05). After the first injection of desipramine, both variables transiently 4 increased (P<0.05) to: 266±4 beats/min and 57±1 mmHg in normoglycaemic rats, and 5 230±12 beats/min and 58±3 mmHg in diabetic rats. These values returned to baseline levels after 6 10 min. The subsequent desipramine treatments did not modify further (P>0.05) these variables.   (Table 2). In contrast, there were no statistically significant 3 differences between the S-R curves in both groups of rats (Table 2).  In view that, in contrast with control normoglycaemic rats (see above), 5-CT was the only 21 agonist mimicking the 5-HT-induced cardiac sympatho-inhibion (Fig. 2D), Fig. 3 further shows  1 μg/kg.min). Interestingly, the sympatho-inhibition to 5-CT, remaining unaffected after 4 1 ml/kg saline (Fig. 4A), was abolished after 1 mg/kg SB 699551 (Fig. 4B), a dose that 5 completely blocks rat 5-HT5A receptors [22,30]. The above i.v. doses of saline or SB 699551 6 were devoid of any effect per se (P>0.05) on the sympathetically-induced tachycardic responses 7 or on baseline heart rate and diastolic blood pressure (data not shown).

8
Expression of the 5-HT1B, 5-HT1D and 5-HT5A receptor subtypes in the stellate ganglion of 9 normoglycaemic and diabetic rats 10 As presented in Fig. 5, the stellate ganglion cells show immunopositive-cells for the 5-HT1B, 11 5-HT1D and 5-HT5A receptor subtypes ( Fig. 5A; white horizontal scale bar: 50 µm). Regardless of 12 the subtype of 5-HT receptor analysed, the reaction was located in the nuclei and cytoplasm of 13 immunopositive-cells. 14 In normoglycaemic (control) animals (Fig. 5B), the 5-HT5A receptor showed a higher 15 intensity of reaction (fluorescence), followed by the 5-HT1B and then by the 5-HT1D receptor. 16 Interestingly, experimental T1D caused a striking and significant increase in the fluorescence 17 intensity of the 5-HT5A receptor, and less marked (but statistically significant) in the 5-HT1B 18 receptor as compared to normoglycaemic rats. On the contrary, the fluorescence intensity of the 19 5-HT1D receptor was slightly -but significantly-reduced in animals with T1D when compared 20 with control animals (Fig. 5B).

General 23
The present study is the first of its kind to establish the prominent relevance of peripheral 5-HT5A 1 receptors modulating the functionality of cardiac sympathetic innervation in a rat model of T1D.

2
To achieve this, we have used the widely known experimental model of T1D induced by STZ 3 administration in rats. This diabetogenic agent causes a syndrome resembling human T1D, 4 characterized by chronic and sustained hyperglycaemia, polyuria, polydipsia, and weight loss, 5 among others, leading to the probable development of both micro and macroangiopathies related 6 to diabetes [31,32]. 7 In addition to the possible therapeutic repercussions derived from this study, the 8 main -and novel-findings of the present investigation in rats with T1D reveal that: (i) 5-HT 9 maintains its cardiac sympatho-inhibitory effects, which were exclusively and potently 10 reproduced by 5-CT (a 5-HT1/5A receptor agonist; [16]); and (ii) prejunctional 5-HT5A receptors 11 were the only receptors involved in the serotonergic cardiac sympatho-inhibition, which are 12 overexpressed in the stellate (sympathetic) ganglion from diabetic rats.

13
Admittedly, the noradrenaline release resulting from stimulation of the cardioaccelerator 14 sympathetic nerves was assessed indirectly by measuring the elicited tachycardic responses, as 15 previously described [15]. Under these conditions, the responses to 5-CT in diabetic rats were 16 considered sympatho-inhibitory (i.e. prejunctional in nature) as this 5-HT1/5A receptor agonist 17 inhibited the tachycardic responses to cardiac sympatho-stimulation without affecting those to 18 exogenous noradrenaline. In fact, our group has already shown in normoglycaemic rats that   23 Since pithed rats were used in this study, any influence of the central nervous system (which can 1 produce baroreflex compensatory mechanisms in response to the effects produced by 5-HT and 2 related agonists) can be excluded. Consequently, the nature of this experimental model helps 3 explain why its resting values for both diastolic blood pressure and heart rate are much lower 4 than those measured in anaesthetised (non-pithed) rats, where the neurogenic sympathetic tone is 5 active [18,19,24,33]. 6 Considering the above, the baseline values of diastolic blood pressure in diabetic rats 7 (48±1 mmHg) were not significantly different from those obtained in normoglycaemic rats 8 (50±2 mmHg). This suggests that the systemic vascular tone and, therefore, the functionality of 9 the systemic vasculature in diabetic rats (in the early stage of T1D) appears to be unaffected by 10 28-day T1D, as previously described [28,29]. In contrast, baseline heart rate values in diabetic 11 rats were significantly lower (219±7 beats/min) than those determined in normoglycaemic rats 12 (249±7 beats/min). This finding coincides with our results previously found in both pithed and 13 conscious rats, with a bradycardic effect that prevails in type 1 diabetic animals [28,29]. These 14 lower heart rate values in diabetic rats may be related to altered heart electrophysiological 15 properties and/or a dysfunction of cardiac β-adrenergic receptors during the diabetic state [34, 16 35]. Notwithstanding this possible cardiac abnormality in diabetic rats, the release of 17 noradrenaline after cardio-selective sympathetic stimulation produced increases in heart rate that 18 did not significantly differ from those in normoglycaemic animals (see Table 2); consequently, 19 the tachycardic responses after electrical stimulation appear to be unaffected in T1D. As noted 20 above, we did not measure the release of noradrenaline after cardio-sympathetic stimulation in 21 both groups. This approach would clearly require other experimental designs that are beyond the 22 scope of the present study. However, tachycardic responses evoked by exogenous noradrenaline 23 in diabetic rats were slightly lower (although only statistically significant at 0.1 and 0.3 µg/kg of 24 noradrenaline) than those found in normoglycaemic rats (see Table 2). On this basis, it could be 1 speculated that: (a) the proven impairment of the cardiac β-adrenergic system in T1D may be due 2 to the attenuation of the tachycardic responses to exogenous noradrenaline and the decrease in  The fact that heart rate and diastolic blood pressure transiently increased after desipramine  , as demonstrated in normoglycaemic rats [22,30].

2
The specific role of cardiac sympatho-inhibitory 5-HT5A receptors in type 1 diabetic rats 3 First, it must be emphasised that the dose used of SB 699551 was high enough to block 4 completely and selectively 5-HT5A receptors in our study, based on two main findings, namely: 5 (i) Its affinity profile, with a pKi=8.2 for 5-HT5A receptors, a pKi ≤ 6.5 for 5-HT1A, 5-HT1B and   these receptors in the present study can be ruled out for a number of reasons, which include that: 2 (i) 5-HT-induced cardiac sympatho-inhibition in normoglycaemic rats, which is resistant to 3 blockade by LY215840 (a selective 5-HT7 receptor antagonist), was abolished by selective 4 antagonists at 5-HT1B, 5-HT1D and 5-HT5A receptors [16,17,22]; and (ii) 5-CT-induced cardiac 5 sympatho-inhibition in diabetic rats was completely blocked by SB 699551 (a selective 5-HT5A 6 receptor antagonist; current results).  [14,20]; and (ii) the heart parasympathetic neurotransmission [42, 43]. 16 The above findings, as a whole, further allow us to hypothesize that the cardiovascular 17 damage caused by chronic hyperglycaemia, primarily evidenced at the cardiac level (as discussed addition, a slight -but significant-increase in the expression of 5-HT1B receptors was found, 16 which could be related to a compensatory mechanism due to the decrease in 5-HT1D receptor 17 expression, but without functional relevance (i.e. with no noticeable sympatho-inhibition  The genesis of CAN is a combination of microvascular damage of the vasa nervorum and 10 neuronal oxidative stress related to an augmented cardiac sympathetic tone, resulting in a 11 sympatho-vagal imbalance and consequent tachycardia, decreased heart rate variability, exercise 12 intolerance, dysrhythmias, silent myocardial impairment or sudden cardiorespiratory death in 13 T1D [4,9,10]. Apart from this neuropathy, other diabetes-associated cardiac pathologies have 14 been related to a sympathetic hyperinnervation, such as atrial fibrillation, congestive heart failure, 15 or ventricular dysfunction [52-54]. 16 Obviously, the alteration of the serotonergic system plays an important role in the Our study suggests that short-term T1D in rats modifies the pharmacological profile of the 5-HT 8 receptors mediating cardiac sympatho-inhibition, so that prejunctional 5-HT5A receptors appear to 9 be the only receptor subtype involved in 5-HT-induced cardiac sympatho-inhibition. Consistent 10 with this suggestion, peripheral 5-HT5A receptors are remarkably overexpressed in type 1 diabetic 11 rats as compared to normoglycaemic rats. These findings provide new evidence that modulation 12 of the peripheral 5-HT system in T1D may help control some cardiac disorders related to 13 sympathetic dysregulation.    The data that support the findings of this study are available from the corresponding author upon 7 reasonable request. 15 Consent for publication: 16 Not applicable 17 Competing interests: 18 The authors declare that they have no competing interests.          control. Δ Heart rate stands for "increase in heart rate".

15
Note that non-significant effects (P>0.05) were produced by the i.v. infusions of saline or 5-CT. 16 Δ Heart rate stands for "increase in heart rate". inhibition of tachycardic responses evoked by cardiac sympathetic stimulation (n=5 each). 20 *P<0.05 vs control. Δ Heart rate stands for "increase in heart rate".